hydrocephalus ii shunt dysfunction and neuroendoscopy...shunt dysfunction and infection ¡infection...
TRANSCRIPT
Hydrocephalus II
Shunt Dysfunction and Neuroendoscopy
Shelly Lwu / Dr. HamiltonMay 18, 2006
Shunt Dysfunctionand Infection
Shunt Dysfunction and Infection
¡ Infectionl Skin breakdown over hardware
¡ Mechanical failurel Undershunting
¡ Separation of shunt components, fractures, migration of hardware
l Overshunting¡ Subdural hematoma
These account for majority of shunt problems
Shunt Dysfunction and InfectionEpidemiology
¡ 25~60% of patientsl 17% in 1st yr after insertion in peds
¡ Higher risk:l Preemiesl Children age <6 mos or weight <3 kg
at time of shunt insertion
Undershunting
UndershuntingEtiology
¡ Blockage within systeml Choroid plexusl Glial adhesionsl Build-up of proteinaceous accretions, blood, cells
(inflammatory or tumor) l Ventricular end most common site
¡ Disconnection, kinking, or breakage of systeml With age, silicone elastomers calcify, break down, &
become more rigid & fragile which may promote subcutaneous attachments
l Barium impregnation may accelerate processl Tube fracture often occurs near clavicle, likely due
to ↑ motion there
UndershuntingEvaluation
¡ Historyl Symptoms of active hydrocephalusl Reason for initial insertion of shuntl Date & reason of last revisionl Type of hardware
¡ Physicall Signs of active hydrocephalusl For children, plot head circumference on graph of normal
curves¡ Before sutures close, head circumferences crossing growth curves
l Swelling along shunt tubing from CSF dissecting along shunt tract
l Ability of shunt reservoir to pump & refill¡ May exacerbate obstruction, esp if shunt is occluded by ependyma due
to overshunting initially¡ In children presenting only w/ N/V, esp those w/ cerebral
palsy & feeding G-tubes, R/O GE reflux
UndershuntingEvaluation
¡ Imagingl Shunt series – plain X-rays
¡ For VP shunt: AP & lateral skull & “low” CXR and/or AXR)¡ R/O disconnection, break, or migration of tip ¡ Disconnected shunt may continue to function by CSF flow thru a
fibrous tract¡ Various hardware may be radioluscent & can mimic disconnection
l U/S¡ Maybe useful in neonates w/ open fontanelles
l CT headl MRIl Radionuclide shunt-o-gram
¡ Assess shunt function using radionuclide, iodinated contrast¡ Shunt tap
l If infection suspected¡ Surgical exploration of shunt
l May be the only means to definitively prove / disprove functioning of various shunt components
l Even when infection not suspected, CSF & removed hardware should be cultured
UndershuntingTreatment
¡ Shunt revision
Pumping the Shunt Reservoir(PS Medical Valve)
¡ To assess flow thru proximal catheter, distal port is depressed first (to occlude run-off into distal catheter), then dome is depressedl Reservoir refills promptly =
shunt patent proximallyl N refill time ~15-30 secs
¡ To assess flow thru distal catheter, proximal port is depressed first, then dome is depressedl Reservoir depresses w/
little resistance = shunt patent distally
¡ Sensitivity 19%, specificity 81% of identifying shunt malfunction
From: Naradzay JFX et al. J Emerg Med. 1999; 17(2):311-322.
Shunt TapIndications
¡ To obtain CSF specimenl To evaluate for shunt infectionl To obtain cells for cytology e.g. in PNET for malignant cellsl To remove blood e.g. in intraventricular hemorrhage
¡ To evaluate shunt functionl Measure pressuresl Contrast studies: proximal injection of contrast (iodinated
or radio-labelled). Distal injection of contrast¡ As temporizing measure to allow function of distally
occluded shunt¡ To inject medication
l Antibiotics for shunt infection or ventriculitisl Chemotherapy agents
¡ For catheters placed within tumor cyst (not a true shunt)l Periodic withdrawal of accumulated fluidl For injection of radioactive liquid (usually phosphorous) for
ablation
Shunt TapTechnique
Step Information Provided
Shave areaPrep w/ providone iodine solution Use 25 gauge butterfly needle (ideally a non-coring needle) - needle should only be introduced into shunt components specifically designed to be tapped
Insert needle into reservoir & look for spontaneous flow into butterfly tubing; measure pressure in manometer
Spontaneous flow = prox end not completely occludedCSF pressure = pressure of ventricular system (should be <15 cm in recumbent position)
Measure pressure w/ distal occluder pressed if present ↑ in pressure = some function of valve & distal shunt
If no spontaneous flow, try to aspirate CSF w/ syringe If CSF easily aspirated, pressure seen by ventricular system may be near 0If no CSF obtained or if difficult to aspirate, prox occlusion
Send CSF for C&S, gram stain, protein, glucose, cell count Check for infection
Fill manometer w/ sterile saline, & occlude proximal (inlet) port
Measure forward transmission pressure (thru valve & peritoneal catheter in presence of shunt w/ proximal occluder) - should be < ventricular pressure
Repeat measurement after injecting 3-5 mL of saline If peritoneal catheter is in loculated compartment, pressure will be ++higher after injection
Radionuclide Shunt-o-gram¡ Position patient, shave hair over reservoir & prep
¡ Tap shunt – insert 25 gauge butterfly needle into reservoirl Measure pressure & drain 2-3 mL of CSF (send 1 mL for C&S)l Inject radio-isotope (for VP shunt in adult, use 1 mCi of 99mTc
pertechnetate in 1 mL of fluid) while occluding distal flow (by compressing valve or occluding ports)
l Flush in isotope w/ remaining CSFl Patients w/ multiple ventricular catheters need to have each injected
to verify patency of that limb
¡ Imagingl Immediately image abdo w/ gamma camera to R/O direct injection
into distal tubingl Image cranium to verify flow into ventricles (proximal patency)l If spontaneous flow into abdo not seen after 10 min, patient sat up &
rescannedl If flow not seen after 10 min, then shunt pumpedl Look for diffusion of isotope within abdo to R/O pseudocyst formation
around catheter
Overshunting
OvershuntingDefinition
¡ Rapid drainage of CSF¡ CSF drainage that occurs when
intraventricular pressure < ventricular valve pressure
¡ Comprises:l True intracranial hypotensionl Slit ventricles & slit ventricle syndrome
OvershuntingEpidemiology
¡ Incidence 5~55%l 10~12% of long-term shunt patients,
within ~6 yrs of initial shunting
¡ Commonly occurs in infants w/ initial shunt insertion at age <6 mos
OvershuntingComplications
¡ Subdural hematomas¡ Craniosynostosis & microcephaly
l Controversial¡ Stenosis or occlusion of sylvian
aqueduct
OvershuntingIntracranial Hypotension
¡ AKA low ICP syndrome¡ Very rare¡ Presentation
l H/A’s postural in nature – worse when upright, relieved w/ recumbency
l Not usually assoc w/, but may occur w/ lethargy, N/V, neuro findings (e.g. diplopia, upgaze palsy)
l May sometimes resemble those of high ICP¡ Etiology: siphoning effect, “true” overshunting¡ CT head: ventricles may be slit-like or N in appearance¡ Sometimes necessary to document drop in ICP (supine
→ upright) for diagnosis¡ Patients may also dev shunt occlusion – may be
difficult to distinguish from slit ventricle syndrome
OvershuntingSlit Ventricles
¡ Totally collapsed lateral ventricles
¡ May be seen on CT head in 3~80% of patients after shunting
¡ Most asymptomatic¡ Patients may
occasionally present w/ symptoms unrelated to shunt (e.g. migraine)
From: Olson S. Pediatr Neurosurg. 2004; 40:264-269.
OvershuntingSlit Ventricle Syndrome¡ AKA non-compliant ventricle syndrome
¡ <12% of shunted patients¡ 6~22% of children w/ radiological slit
ventricles & H/A’s
¡ Triad:l Intermittent clinical features of shunt
obstruction w/ distinct asymptomatic intervalsl Slit-like appearance of ventricles on CT scanl Slow refill of shunt reservoir
OvershuntingSlit Ventricle Syndrome¡ Pathophysiology: small ventricles predispose to
catheter obstruction, pressure then rises & only when ventricles marginally dilate does catheter begin to function again
l Theories – likely multifactorial¡ Ventricular pressure intimately related to ICP & when CSF
pressure drops uncoupling occurs → ↑venous congestion & ↑ brain elastance
¡ ↑ pressure w/ subependymal flow can cause subependymal & periventricular gliosis w/ ↑ ventricular wall stiffness
l Intraventricular pressure would need to be higher than usual to obtain ventricular dilatation
l (Law of Laplace P=2T/R: pressure required to expand a large container < pressure required to expand small container)
l Matsumoto et al. (1986) – animal models¡ Low pressure valves in neonates lead to overshunting w/
radiological slit ventricles, development of microcephaly & craniosynostosis
l Predisposes to ventricular catheter obstruction & prevents ventricles from expanding in response
OvershuntingEvaluation
¡ Clinical examl Deep sunken fontanellel Overriding parietal bonesl Rapid decline in head circumference to microcephalic
range l Valve slow to refill after compression
¡ Monitoring CSF pressure l Lumbar drainl Shunt tap – measure pressures w/ postural changes
¡ –ve pressure when uprightl Pressure spikes during sleep
¡ CT headl Slit like ventriclesl May show evidence of transpendymal CSF flowl Possible SDH / ICH
¡ Shunt-o-gram
OvershuntingTreatment of Slit Ventricle Syndrome
¡ Try to categorize patientl If possible, then implement specific txl Otherwise: tx empirically as intracranial
hypotension, then move onto other methods for tx failure
¡ Problems related to overshunting may be reduced by utilizing LP shunts for communicating hydrocephalus & reserving ventricular shunts for obstructive HCP
¡ VP shunts may also be more likely to overdrain than VA shunts b/c longer tubing resulting in greater siphoning effect
OvershuntingTreatment of Slit Ventricle Syndrome
¡ Intracranial hypotensionl Postural H/A usually self-limitedl Symptoms persistent after >3
days bed-rest & analgesics¡ Trial w/ tight abdo binder¡ Valve should be checked for
proper closing pressure: if low, replace w/ higher pressure valve; if not low, antisiphon device +/- high pressure valve
¡ Patients w/ long-standing overshunting may not tolerate efforts to return intraventricular pressure to N levels
¡ Asymptomatic slit ventriclesl Prophylactic upgrading to
higher pressure valve or insertion of antisiphon device now largely abandoned
l May be appropriate at time of shunt revision when done for other reasons
¡ Slit ventricle syndromel Patients actually suffering from
intermittent high pressurel Total shunt malfunction: revise shuntl Intermittent occlusion:
¡ If symptoms occur early after shunt insertion / revision, initial expectant management may be indicated since symptoms will spontaneously resolve in many
¡ Revision of proximal shunt (may be difficult due to small size of ventricles): follow existing tract & insert longer / shorter length of tubing based on pre-op imaging studies vs. insertion of 2nd
ventricular catheter (leave 1st one in place) / LP shunt
¡ Upgrade valve / programmable valve
¡ Antisiphon device¡ Subtemporal decompression
sometimes w/ dural incision →dilatation of temporal horns (evidence for ↑ ICP) in most, but not all cases
¡ Calvarial expansion¡ 3rd ventriculostomy
OvershuntingTreatment of Slit Ventricle Syndrome
From: Olson S. Pediatr Neurosurg. 2004; 40:264-269.
Infection
InfectionRate
¡ Majority of data from retrospective reviews using data collected before 1990l 18~22% of shunted patientsl 5~6% per procedure
l 6.2% in 1st post-op month, 7.4% overalll 9% by 6 months, 19% by 10 yrs
¡ Education Committee of International Society of Pediatric Neurosurgeons sponsored cooperative study (1994) – 38 centers, 773 patients, >1 yr follow-upl 6.5% of patients after 1 yr
InfectionTime to Infection
¡ Majority soon after placement of shuntl Casey et al. (1997)
¡ 92% occur within 3 months of shunt placement
InfectionRisk Factors
¡ Age – neonates & very young childrenl Casey et al. (1997)
¡ ↑ infection rate in age <6 months vs. older (19% vs. 7%)
l Age-related changes in density & identity of bacteria populations on skin, ↑ susceptibility to pathogens due to relative immune deficiency in neonates (less IgG levels in age <6 months)
¡ Duration of shunt surgery¡ Previous shunt failure
¡ Possible assoc risk factors:l Reason for shunt
placementl Type of shuntl Educational level of
surgeonl Presence of spinal
dysraphism – few studies w/ enough statistical power
¡ Ammirati & Raimondi et al. (1987): subgroup analyses
l ↑ infection rate in myelomeningocele patients shunted in 1st
wk of life vs. age >2 wks (50% vs. 24%)
¡ Clinically stable children might benefit from delay in shunt placement
InfectionPresentation¡ Varies w/ age
¡ Infants: ↑ irritability, apnea & bradycardia in severe cares
¡ General:l H/Al Lethargyl N/Vl Feverl Meningismusl Photopobial Gait disturbancesl Seizuresl Visual disturbances (upward
gaze palsy & papilledema)l Abdo pain, abdo fluid
collection / pseudocystl Erythema or edema along
shunt tubing
¡ Varies w/ organism
¡ Gram –ve bacilli – E. coli: acute presentation w/ severe abdo pain, septicemia
¡ S. epidermidis: more indolent¡ S. aureus: usually assoc w/
erythema along shunt tract
¡ Ventricular-vascular shunts:l Subacute bacterial
endocarditisl Shunt nephritis (immune
complex deposition in renal glomeruli) – hematuria + proteinuria
InfectionEvaluation
¡ History & physical
¡ Imagingl X-ray shunt series – R/O disconnection in shunt tubing,
movement of distal catheter out of peritoneal space w/ growth of child
l CT – ependymal enhancement characteristic of ventriculitis¡ Difficult in children w/ slit ventricles or unusual baseline ventricular
anatomyl U/S for neonatesl Abdo U/S – R/O fluid collection
¡ Shunt tapl Opening pressure, functionl CSF for glc, protein, cell count, gram stain, C&S
¡ ↓ glc, ↑ protein, ↑ cell count suggest bacterial infection¡ Generally, C&S +ve in <50% tested
InfectionPrevention
¡ Langley et al. (1993) – meta-analysis, 12 studiesl Peri-op abx use reduced infection rate by ~50% (CI
95%)¡ Haines & Walters (1994) – meta-analysis, 8
studiesl Same result
¡ Studies included in meta-analyses differed wrt specific type, duration, dosage of abx used, as well as timing of 1st dose
¡ Little evidence to suggest that abx prophylaxis before dental procedures & use of abx-impregnated silastic shunts reduce infection rates
InfectionOrganisms
¡ Infection occurs via:l Bloodstreaml Along shunt tubing from abdo source (generally assoc w/
bowel perforation)l Contamination of shunt material w/ skin flora @ time
of surgery
¡ Most common: typical skin flora¡ In most series: S. epidermidis > S. aureus (2:1)
l Livni et al. (2004) ¡ S. epidermidis secretes mucoid slime that enhances its ability to
adhere to foreign bodies¡ Lower adherence rate for silicone vs. teflon
¡ Gram –ve bacteria: E. coli, Proteus, Klebsiellal From intestinal perforation
¡ Anaerobic diphtheroids, e.g. propionibacterium, can cause delayed infectionsl Difficult to assess & tx as C&S may remain –ve for >1 wk
¡ Fungal infections rare
InfectionOrganisms
¡ Infection = +ve C&S from CSF or from shunt hardware
¡ In most instances, only shunt hardware is C&S +ve while CSF remains –vel Vanaclocha et al. (1996)
¡ C&S positivity hardware vs. CSF (59% vs. 9%)
l Suggests bacteria & other microorganisms favor adhesion to foreign materials over CSF
InfectionTreatment
¡ Surgical removal of infected shunt or shunt externalization
¡ New shunt placed then or delayed until after course of abx
¡ May need EVD, lumbar drain, or intermittent LP’s or ventricular taps as temporizing measure
¡ Abx course until CSF C&S –ve for minimum 72 hrs
¡ Drainage of abdo pseudocyst / abscess / wound may also be necessary
¡ IV vancomycin often used initially until sensitivities available
InfectionTreatment
¡ Abx alone less effective than abx + surgeryl Walters et al. (1984) – retrospective review
¡ 14% vs. 60%l Frame & McLaurin (1984) – RCT
¡ Removal of shunt + abx + interim EVD or ventricular taps vs. immediate surgical replacement of shunt + abx vs. abx only (IV & intraventricular abx given to all)
¡ Higher cure rates @ 48 hrs, 1 & 4 months for surgery patients: 100% vs. 90% vs. 30%
¡ Longer hospital stays for abx-only patients: 47 vs. 33 vs. 25 days
InfectionTreatment: AntibioticsEmpiric abx¡ IV vancomycin initially
(CSF penetration ~18% conc of serum)
¡ Consider adding PO rifampin for increased coverage (10 mg/kg/day PO q12h)
¡ May change to IV nafcillin (unless penicillin allergic or MRSA +ve)
¡ Intraventricular injection of preservative-free abx
Abx for specific organisms¡ S. aureus & S. epidermidis
l If sensitive (MIC <1.0 µg/mL): IT gent + (IV nafcillin / cefazolin / cephalothin / cephapirin)
l If resistant to nafcillin (i.e. MRSA) / cephalothin / cephapirin: PO rifampin + PO trimethoprim + IV & IT vancomycin
¡ Enterococcus: IV/IT ampicillin + IT gent (if intravascular shunt: add IV gent)
¡ Other streptococci: either antistreptococcal or above enteroccal regimen
¡ Aerobic GNB: based on susceptibilities; both beta-lactam & antipseudomonal aminoglycosides IV & IT indicated
¡ Corynebacterium sp. & Proprionibacterium sp. (diphtheroids)
l If penicillin sensitive: use enterococcal regimen above
l If penicillin resistant: IV + IT vancomycin
InfectionTreatment: Antibiotics
From: Naradzay JFX et al. J Emerg Med. 1999; 17(2):311-322.
InfectionAssociated Outcomes
¡ Extended hospital stays¡ ↑ long-term mortality risk¡ ↑ seizure risk ¡ Delayed developmental milestones¡ Lower IQ & poor school
performance
Neuroendoscopy
NeuroendoscopyHistory¡ 1910 – L’Espinase – attempted fulguration of choroid plexus in 2 infants w/
hydrocephalus with cystoscope – 1 patient died post-op¡ 1922 – Dandy – similar
¡ 1923 – Fay & Grant – visualized & photographed interior of ventricles of child w/ hydrocephalus w/ cystoscope
¡ 1923 – Mixter – 1st successful 3rd ventriculostomy
¡ 1943 – Putnam – endoscopic choroid plexectomies by cauterization – high failure & peri-op mortality rates
¡ 1970 – Scarff – similar
¡ Decline in neuroendoscopy w/ advent of ventricular shunts & development of microsurgery
¡ Rediscovery of neuroendoscopy w/ advances in technology in 1970’s
¡ 1990 – Jones – 50% shunt-free success rate for endoscopic 3rd
ventriculostomy – improved to 60% in subsequent series
NeuroendoscopyTechnology
¡ Ventricular cannula
¡ Endoscope:l Rod-lens scope
(rigid)¡ Clearer images
l Fiberscope (flexible)
¡ Other ports:l Electrocauteryl Irrigation
¡ RL or NS
¡ Camera¡ Video monitor¡ Light source
l Halogen, mercury vapor, xenon
NeuroendoscopyUses
¡ Hydrocephalusl Obstructive
hydrocephalus from primary aqueductal stenosis or compressive periaqueductal mass lesions
l Septum pellucidotomy or septostomy for isolated lateral ventricles
l Fenestration of loculated ventricles
l Marsupialization & fenestration of intracranial cysts
l Aqueductoplasty
¡ Neurooncologyl Biopsy & resection of
intraventricular tumorsl Resection of colloid
cystsl Endonasal
transsphenoidal hypophysectomy
¡ Spine surgeryl Thoracoscopic
sympathecotmyl Discectomyl Lumbar laminotomyl Resection of tumors &
cysts¡ Craniosynostosis
Endoscopic Third Ventriculostomy
¡ For tx of obstructive hydrocephalus caused by primary aqueductal stenosis or compressive periaqueductal mass lesions
¡ More physiologic tx of obstructive hydrocephalus by allowing egress of ventricular CSF directly into subarachnoid space, bypassing downstream stenosisl Opening made in the floor of 3rd ventricle
¡ Alternative to VP shunt which is assoc w/ frequent & multiple complicationsl Opportunity for patient to have shunt-free existence
Endoscopic Third VentriculostomyPatient Selection
¡ Symptoms & signs of hydrocephalus
¡ Features on MRIl Enlarged lateral & 3rd ventricles, w/ N
or small 4th ventriclel Midsagittal section demonstrating
adequate space between basilar artery & clivus under floor of 3rd ventricle
Endoscopic Third VentriculostomyAnatomy
From: Li KW et al. Neurosurg Focus. 2005; 19(6):E1.
Endoscopic Third VentriculostomyAnatomy
From: Jallo GI et al. Neurosurg Focus. 2005; 19(6):E11.
Endoscopic Third VentriculostomyAnatomy
From: Jallo GI et al. Neurosurg Focus. 2005; 19(6):E11.
Endoscopic Third VentriculostomyTechnique
¡ Burr hole at or just anterior to coronal suture, 2.5-3 cm lateral to midline
¡ Open dura in cruciate fashion, coagulate edges¡ No. 14 Fr catheter used to cannulate lateral ventricle
l Remove stylet¡ Pass endoscope thru sheath to visualize lateral
ventricle¡ Identify Foramen of Monro & navigate scope into 3rd
ventriclel Identify mamillary bodies & infundibular recess in floorl Sometimes basilar artery visible
¡ Puncture floor of 3rd ventricle & dilate opening¡ On completion, remove scope & sheath¡ Gelfoam plug in burr hole¡ Close galea & skin
Endoscopic Third VentriculostomyPost-op Care
¡ Observation in ICU x 1 day
¡ MRIl CINE – CSF flow thru opening in floor
of 3rd ventriclel Ax T2WI (“Poor man’s CINE”) – flow
void in floor of 3rd ventricle
Endoscopic Third VentriculostomyOutcome
¡ Overall success rate 50~90%
¡ Most failures occur soon after procedurel Reclosure of ventriculostomy ~22%
¡ Longer follow-up studies necessary
Endoscopic Third VentriculostomyPossible Complications¡ Incidence 0~20%
¡ Bleedingl SAH - injury to basilar arteryl ICH l IVH - bleeding from choroid plexusl SDH
¡ Injury to surrounding structuresl Cranial nerve palsy – CN III, VIl Fornix, caudate, thalamus, thalamostriate venous
complexl Hypothalamic / pituitary dysfunction
¡ Typically manifests as DI ¡ Cardiac arrhythmias or resp arrest from manipulation or
irritation of hypothalamus¡ Infection
¡ Mortality ~1%
References¡ Amini A and RH Schmidt. “Endoscopic third ventriculostomy in
adult patients.” Neurosurg Focus. 2005; 19(6):E9. ¡ Garton HJL and JH Piatt. “Hydrocephalus.” Pediatr Clin N Am.
2004; 51:305-325.¡ Greenberg MS. Handbook of Neurosurgery. New York: Thieme,
2001. Pp. 185-191.¡ Jallo GI et al. “Endoscopic third ventriculostomy.” Neurosurg
Focus. 2005; 19(6):E11.¡ Li KW et al. “Neuroendoscopy: past, present, and future.”
Neurosurg Focus. 2005; 19(6):E1.¡ Livni G et al. “In vitro bacterial adherence to ventriculoperitoneal
shunts.” Pediatr Neurosurg. 2004; 40:64-69.¡ Naradzay JFX et al. “Cerebral ventricular shunts.” J Emerg Med.
1999; 17(2):311-322.¡ Olson S. “The problematic slit ventricle syndrome.” Pediatr
Neurosurg. 2004; 40:264-269.¡ Winn KH. Yeomans Neurological Surgery, 5th ed. New York:
Saunders, 2004. Pp. 3419-3432.